Human Reproduction Vol.22, No.4 pp. 980–988, 2007 doi:10.

1093/humrep/del484
Advance Access publication January 4, 2007

Milder ovarian stimulation for in-vitro fertilization

reduces aneuploidy in the human preimplantation
embryo: a randomized controlled trial
Esther B.Baart1,2,6, Elena Martini2, Marinus J.Eijkemans3, Diane Van Opstal4, Nicole
G.M.Beckers2, Arie Verhoeff5, Nicolas S.Macklon1 and Bart C.J.M.Fauser1,2
1
Department of Reproductive Medicine and Gynaecology, University Medical Center, Utrecht, The Netherlands, 2Division of
Reproductive Medicine, Department of Obstetrics and Gynaecology, 3Department of Public Health, 4Department of Clinical Genetics,

Downloaded from https://academic.oup.com/humrep/article-abstract/22/4/980/697233 by guest on 14 June 2019

Erasmus Medical Center, Rotterdam, The Netherlands and 5Department of Obstetrics and Gynaecology, Medical Center Rijnmond Zuid,
Rotterdam, The Netherlands
6
To whom correspondence should be addressed at: Department of Reproductive Medicine and Gynecology, University Medical Center,
Heidelberglaan 100, 3584 CX Utrecht, The Netherlands. Tel.: þ31 30 250 1443; Fax: þ31 30 250 5433; E-mail: e.b.baart@umcutrecht.nl

BACKGROUND: To test whether ovarian stimulation for in-vitro fertilization (IVF) affects oocyte quality and thus
chromosome segregation behaviour during meiosis and early embryo development, preimplantation genetic screening
of embryos was employed in a prospective, randomized controlled trial, comparing two ovarian stimulation regimens.
METHODS: Infertile patients under 38 years of age were randomly assigned to undergo a mild stimulation regimen
using gonadotrophin-releasing hormone (GnRH) antagonist co-treatment (67 patients), which does not disrupt
secondary follicle recruitment, or a conventional high-dose exogenous gonadotrophin regimen and GnRH agonist
co-treatment (44 patients). Following IVF, embryos were biopsied at the eight-cell stage and the copy number of 10
chromosomes was analysed in 1 or 2 blastomeres. RESULTS: The study was terminated prematurely, after an
unplanned interim analysis (which included 61% of the planned number of patients) found a lower embryo aneuploidy
rate following mild stimulation. Compared with conventional stimulation, significantly fewer oocytes and embryos
were obtained following mild stimulation (P < 0.01 and <0.05, respectively). Consequently, both regimens generated
on average a similar number (1.8) of chromosomally normal embryos. Differences in rates of mosaic embryos suggest
an effect of ovarian stimulation on mitotic segregation errors. CONCLUSIONS: Future ovarian stimulation strategies
should avoid maximizing oocyte yield, but aim at generating a sufficient number of chromosomally normal embryos
by reduced interference with ovarian physiology.

Key words: aneuploidy/embryo quality/in-vitro fertilization/ovarian stimulation/preimplantation genetic screening

Introduction
Human reproduction is a relatively inefficient process (Norwitz
et al., 2001). The chance of achieving a spontaneous pregnancy
after timed intercourse is 20 – 30% (Evers, 2002; Taylor, 2003),
significantly lower than 70% in the rhesus monkey (Ghosh
et al., 1997), 80% in captive baboons (Stevens, 1997) or 90%
in rodents and rabbits (Foote and Carney, 1988). Moreover,
up to 30% of early human embryos fail to develop into
viable fetuses (Wilcox et al., 1988), largely due to chromo-
somal abnormalities (Boué et al., 1975; Vorsanova et al.,
2005). The incidence of embryo aneuploidy increases with
maternal age (Hassold and Hunt, 2001).
In-vitro fertilization (IVF) is the major treatment strategy for
infertility, employing complex and costly ovarian stimulation
protocols to generate multiple embryos (Fauser et al., 2005;
Macklon et al., 2006). After ovarian stimulation and IVF, the
best quality embryos are selected for transfer into the uterine
cavity. Although embryo morphology is widely used to evalu-
ate embryo quality, this subjective method provides only
limited information concerning the chromosomal constitution
(Munné, 2006). The introduction of fluorescence in-situ
hybridization (FISH) on interphase nuclei allowed the screen-
ing of embryos for chromosomal aneuploidies, a procedure
referred to as preimplantation genetic screening (PGS) (Thorn-
hill et al., 2005). Clinically, PGS is being advocated for older
women (Munné et al., 2003; Staessen et al., 2004) and for
patients with recurrent spontaneous abortion or repeated
implantation failure (Gianaroli et al., 2003; Pehlivan et al.,
2003; Platteau et al., 2005). High rates of aneuploidy have
been reported in these women. Moreover, in studies where
the entire embryo was analysed, a high incidence of chromoso-
mal mosaicism has been observed (Delhanty et al., 1993;
Bielanska et al., 2002). The frequent occurrence of mosaicism,

980 # The Author 2007. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology.
All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

resulting from mitotic segregation errors (Delhanty, 1997), is
also reflected in the high incidence of discordant FISH
results when two blastomeres are analysed by PGS (Baart
et al., 2004b, 2006).
The mechanisms underlying aneuploidy are still poorly
understood. However, recent observations suggest that inac-
curacies of the chromosome segregation machinery in
oocytes are often involved, and this process is influenced by
maternal age (Hassold and Hunt, 2001; Champion and
Hawley, 2002). Preliminary observations suggest that aneu-
ploidy in embryos may also be affected by ovarian stimulation
regimens employed in IVF (Munné et al., 1997; Katz-Jaffe
et al., 2005). Conventional IVF regimens routinely use a
gonadotrophin-releasing hormone (GnRH) agonist long
protocol co-treatment to prevent a premature luteinizing
hormone (LH) rise. Down-regulation of pituitary function
takes around 2 weeks, after which high doses of exogenous
FSH are administered to induce multiple follicle growth. The
recent availability of GnRH antagonists has enabled the
development of milder approaches in ovarian stimulation. To
prevent an LH rise, GnRH antagonist administration can be
limited to the mid-to-late follicular phase (Fauser and
Devroey, 2005), allowing the endogenous inter-cycle, FSH
rise to be utilized for follicle stimulation. Cyclic follicle
recruitment and initial stages of dominant follicle selection
can proceed within the natural cycle and the use of exogenous
FSH for inducing multiple follicle development can be
restricted to the mid-late follicular phase (Fauser and
Van Heusden, 1997; Fauser et al., 1999; Hohmann et al., 2003).
To test whether the conventional ovarian stimulation proto-
col and a mild stimulation approach differentially affect the
competence of oocytes and embryos for proper chromosome
segregation, PGS was employed in a prospective randomized
controlled trial in a group of IVF patients younger than 38
years of age.
Figure 1. Schematic representation of the two ovarian stimulation protocols for IVF and laboratory procedures for preimplantation genetic
screening (PGS).
Ovarian stimulation affects oocyte quality
Materials and methods
Study design
All patients were recruited from the outpatient clinic at the
Erasmus Medical Center and the Medical Center ‘Rijnmond
Zuid’ from December 2002 to August 2005. Patients were ran-
domly assigned to undergo a mild ovarian stimulation regimen
using GnRH antagonist co-treatment or a conventional high-dose
gonadotrophin regimen and GnRH agonist long protocol
co-treatment. A schematic representation of the study is outlined
in Figures 1 and 2. A population of infertile couples was targeted,
who were not at an a priori increased risk for chromosomally
abnormal embryos. Only women below 38 years of age, with a
regular indication for IVF and with a partner with a sperm count
.5 million progressively motile sperm per millilitre (prior to capa-
Downloaded from https://academic.oup.com/humrep/article-abstract/22/4/980/697233 by guest on 14 June 2019
citation) were invited to participate. Additional inclusion criteria
were: history of regular menstrual cycles (ranging from 25 to
35 days), body mass index between 19 and 29 kg m22, no
known chromosomal abnormalities, no relevant systemic disease
or uterine and ovarian abnormalities, no history of recurrent mis-
carriage, and no previous IVF cycles not resulting in an embryo
transfer. Couples could participate in the study for one cycle
only. Prior to commencing the study, ethical approval was received
from the Dutch Central Committee on Research Involving Human
Subjects (CCMO) and the local institutional Ethics Committee.
Written informed consent was obtained from each couple.
A higher cancellation rate before oocyte retrieval and less
embryos was expected following mild ovarian stimulation
(Hohmann et al., 2003). Therefore, randomization to one of the
two treatment groups was performed according to a computer-
generated randomization schedule in a ratio of 4 : 6 (conventional
group: mild group; see statistical paragraph), assigned via numbered
sealed envelopes. After the patient agreed to participate, the next
available numbered envelope on entry into the study was opened
by the treating physician during the preparatory IVF consultation.
Blood samples were drawn from each patient on cycle Day 3 or
4 before the start of stimulation, to assess baseline FSH and
inhibin B levels.
981
E.B.Baart et al.
Figure 2. Trial profile and flow of patients. Ovarian hyperstimulation syndrome (OHSS), stimulation was discontinued because of signs of
an imminent OHSS; low response, only one growing follicle observed at ultrasound; ICSI, Intracytoplasmic sperm injection necessary due to
unexpected poor semen quality on the day of oocyte retrieval.
Multifollicular ovarian stimulation
Patients randomized to undergo conventional ovarian stimulation
were treated for at least 2 weeks with the GnRH agonist Triptorelin
(Decapeptylw, Ferring BV, Hoofddorp, The Netherlands) 0.1 mg per
day s.c., starting 1 week before the expected menses. Following pituitary
down regulation, patients received a fixed daily dose of 225 IU s.c.
recombinant FSH (Puregonw, NV Organon, Oss, The Netherlands).
Preferably, FSH treatment was started on Mondays or Tuesdays
to reduce chances for a biopsy procedure on weekends. Patients
randomized to the mild stimulation protocol were treated with a
fixed dose of 150 IU s.c. recombinant FSH (Puregonw) starting
on cycle Day 5. GnRH antagonist co-treatment (Orgalutranw, NV
Organon) at 0.25 mg per day s.c. was initiated on the day the
leading follicle reached a diameter of 14 mm (Hohmann et al.,
2003). To induce final oocyte maturation, a single dose of
10 000 IU s.c. hCG (Pregnylw, NV Organon Oss, The Netherlands)
was administrated as soon as the leading follicle had reached
a diameter of 18 mm and at least one additional follicle had
reached a diameter of 15 mm. Oocyte retrieval was carried out
35 h after hCG injection by transvaginal ultrasound-guided punc-
ture of follicles.
982
Downloaded from https://academic.oup.com/humrep/article-abstract/22/4/980/697233 by guest on 14 June 2019
In-vitro fertilization, embryo culture and biopsy
After oocyte retrieval, IVF and embryo culture were performed as
described previously (Huisman et al., 2000; Hohmann et al., 2003).
On day 3 after oocyte retrieval, embryos resulting from normally fer-
tilized oocytes (as evidenced by two visible pronuclei) were scored
according to previously published morphological criteria (Hohmann
et al., 2003), blinded to the stimulation protocol. These included
cell number, regularity of blastomeres, fragmentation and morpho-
logical aspects including granulation. Normal morphology was
defined as embryos with timely development, ,20% fragmentation,
about equal sized blastomeres and small or no irregularities observed
in the cytoplasm. Biopsy was performed on embryos with more than
five blastomeres. Two cells were removed unless the embryo consisted
of only six blastomeres. Embryo biopsy and fixation of biopsied cells
were performed as described elsewhere (Baart et al., 2004a).
FISH analysis and diagnosis
FISH analysis was performed to determine the copy number of nine
chromosome pairs (1, 7, 13, 15, 16, 18, 21, 22, X and Y), as previously
described (Baart et al., 2004, 2004b). FISH results were interpreted by

two independent observers, blinded to the stimulation protocol. For
enumeration of the signals on single blastomere nuclei, we used pre-
viously published scoring criteria (Munné et al., 1998). A nucleus
was considered normal if it showed the normal (diploid) amount of
signals for the chromosomes investigated and abnormal if one or
more of the chromosomes investigated showed an increased or
decreased number of signals. In case two cells were available,
embryos were classified as normal (both nuclei normal FISH
results), uniformly abnormal (both nuclei showing the same abnorm-
ality) or mosaic (one normal and one abnormal nucleus or two
abnormal nuclei with each nucleus showing different chromosome
abnormalities). No more than two normal embryos were transferred
to the patient.
As a result of chromosomal mosaicism, the definition of an abnor-
mal embryo is different if one cell is available for analysis when com-
pared with two available cells. Also, embryos where only one cell
could be biopsied differ developmentally from embryos where a
two-cell biopsy was possible. To obtain uniformity for statistical
analysis, we used two approaches. First, all embryos were classified
in retrospect as either normal or abnormal on the basis of the FISH
results obtained from the first biopsied blastomere, even if two cells
were available. Second, the analysis was repeated for only those
embryos with a PGS diagnosis based on two cells.
Outcome measures
Primary outcome measures were ovarian response, as assessed by the
number of oocytes obtained and the proportion of chromosomally
abnormal embryos per patient. This was expressed as the ratio of
abnormal embryos on the number of embryos diagnosed per patient.
Secondary outcome measures were the proportion of fertilized
oocytes, the proportion of embryos with normal morphology and the
proportion of embryos biopsied and diagnosed. All proportions were
first calculated per patient and then averaged for each treatment
group. As women were randomly assigned to two different ovarian
stimulation protocols to detect possible differences in chromosomal
abnormality rates of embryos generated, this is the correct unit of
statistical analysis.
Statistical analysis
Before commencing the study, the sample size was determined. We
assumed a reduction in the aneuploidy rate from 30% after conven-
tional ovarian stimulation to 20% after mild ovarian stimulation. We
calculated that 293 embryos in each group would achieve an 80%
power to detect this 10% difference at an alpha level of 0.05 with
the use of a two-sided t-test. With an expected average of six
embryos following conventional and four following mild ovarian
stimulation and an expected drop-out rate of one-third of the
patients from each group, the total number of subjects to be included
was 73 patients in the conventional group and 109 patients in the mild
Table I. Baseline characteristics for IVF patients in the two different treatment groups for ovarian stimulation
Female age (years)
FSH level on cycle Day 3 or 4 (IU l21)
Inhibin B level on cycle Day 3 or 4 (ng l21)
No. of previous IVF cycles, n (%)
0 32 (73)
1 3 (7)
2 6 (14)
3 3 (7)
Data are expressed as median values and range, unless otherwise stated.
Ovarian stimulation affects oocyte quality
group. However, due to slow patient inclusion and an increasing
concern regarding the safety of a two-cell biopsy with respect to the
implantation potential of the embryo (Cohen and Munné, 2005),
an unplanned interim analysis was performed after the inclusion of
111 patients. The proportion of chromosomally abnormal embryos
per patient was found to be significantly reduced after mild
ovarian stimulation [P ¼ 0.02, which is below the Pocock critical
bound of 0.0354 for a single interim analysis after 61% (111 of 181)
of patients had been included (Pocock, 1977)] and the study was
terminated.
A x2 test was used to test for differences between the two groups in
the percentage of patients with oocyte retrieval and embryo biopsy. A
t-test was used to test for differences in continuous variables and par-
ameters that were, per patient, averaged over oocytes or over embryos,
Downloaded from https://academic.oup.com/humrep/article-abstract/22/4/980/697233 by guest on 14 June 2019
e.g. the average morphology score of the embryos or the percentages
of abnormal embryos. Pearson’s correlation coefficient was used to
test for assocation between parameters of ovarian response and the
proportion of abnormal embryos. To see whether these associations
differed between the two groups, a test for interaction in analysis
of variance (ANOVA) was used. P-values ,0.05 were considered
statistically significant, except for the proportion of chromosomally
abnormal embryos per patient, the primary outcome measure, where
P , 0.0354 was considered significant, according to Pocock’s
method for interim analysis (see above).
Results
Patient and study characteristics
One-hundred and eleven patients were included. Initial screen-
ing characteristics (median and range) for both groups are
presented in Table I. There were no significant differences
between the groups in demographic variables or initial screen-
ing parameters. In 73 (66%) patients, IVF treatment resulted in
the availability of embryos for PGS. Reasons for patient
drop-out and exclusion from analysis are given in Figure 2.
Table II presents the number of oocytes retrieved and success-
fully fertilized and the biopsy results. The number and results of
embryos successfully analysed on one or two blastomeres are
given. To obtain uniformity for statistical analysis, the embryos
were classified in retrospect as either normal or abnormal on
the basis of the FISH results obtained from the first biopsied
blastomere, even if two cells were available. This resulted in
61/159 (38%) chromosomally normal embryos in the conven-
tional stimulation group and 71/143 (50%) normal embryos in
the mild stimulation group. The proportion of normal embryos
was subsequently calculated per patient, as were the other
primary and secondary outcome measures.
Conventional stimulation (n ¼ 44) Mild stimulation (n ¼ 67)
34.1 (28– 37) 33.2 (22– 37)
8.1 (4.4– 13.8) 7.6 (5.5–18.4)
86 (2–1056) 88 (15– 593)
55 (82)
3 (4)
4 (6)
5 (7)
983
E.B.Baart et al.
Table II. Number of embryos biopsied and analysed by Fluorescent in-situ hybridization (FISH) on one or two blastomeres after conventional or mild
stimulation
No. of oocytes obtained
No. of embryos (2pn) obtained
No. of embryos suitable for biopsy
No. of embryos diagnosed
No. of embryos diagnosed based on two cells
Normal
Abnormal
Abnormal/normal mosaic
Abnormal/abnormal mosaic
No. of embryos diagnosed based on one cell
Normal
Abnormal
Values between parentheses are percentages.
Chromosomal competency of embryos correlates with
ovarian response after mild stimulation
The distribution of the number of oocytes retrieved per patient
was different following conventional and mild ovarian stimu-
lation, with skewing of the curve following mild stimulation
towards fewer oocytes (Figure 3a and b). For each stimulation
protocol, differences in the proportion of abnormal embryos
based on one-cell diagnosis were correlated to ovarian response
per patient (Figure 3c and d). Within the mild group, a signifi-
cant positive correlation (Pearson correlation ¼ 0.4;
P ¼ 0.006) was observed between the number of oocytes
obtained and the proportion of abnormal embryos. In the
conventional stimulation group, no correlation was observed
(Pearson correlation ¼ 20.08; P ¼ 0.679). The distribution
found after mild stimulation was significantly different from
the one found after conventional stimulation (P ¼ 0.016; test
for interaction in ANOVA).
Mild ovarian stimulation results in a reduced proportion of
abnormal and mosaic embryos
Table III summarizes outcome measures and clinical results
per patient. Although more oocytes were obtained per patient
following conventional ovarian stimulation (12.1 versus 8.2,
P ¼ 0.001), no differences were observed in fertilization
rates or percentage of embryos biopsied and diagnosed
between the groups. The proportion of embryos with normal
morphology was higher after mild, when compared with
conventional ovarian, stimulation (51 versus 35%; P ¼ 0.04).
On the basis of the first cell biopsied, the proportion of
chromosomally abnormal embryos per patient was significantly
decreased following mild stimulation (Table III). The percentage
of abnormal embryos relative to the number of embryos
diagnosed was 45% following mild stimulation (40 patients)
compared with 63% following conventional stimulation
(33 patients; P ¼ 0.02). Mild stimulation resulted in signifi-
cantly less oocytes and embryos, but there was no difference
between the two study groups in the average number of chromo-
somally normal embryos (1.8) obtained per patient (Figure 4).
By analysing the group of embryos in which two cells were
available for diagnosis, insight into chromosomal mosaicism
could be obtained (Table III). In this group, the diagnosis could
984
Conventional stimulation Mild stimulation
484 459
271 260
184 (68) 157 (60)
159 (86) 143 (91)
98 (62) 96 (67)
27 (28) 37 (39)
12 (12) 14 (15)
32 (33) 20 (21)
27 (28) 25 (26)
61 (38) 47 (33)
20 (33) 16 (34)
41 (67) 31 (66)
Downloaded from https://academic.oup.com/humrep/article-abstract/22/4/980/697233 by guest on 14 June 2019
be normal, abnormal or mosaic. Overall abnormality rates (abnor-
mal and mosaic embryos) were 55% following mild (38 patients)
and 73% following conventional ovarian stimulation (30 patients;
P ¼ 0.046), confirming the difference in abnormality rates
observed after single-cell diagnosis. However, the proportion of
mosaic embryos per patient was more significantly increased
following conventional ovarian stimulation (65 versus 37%;
P ¼ 0.004). This observation indicates that the increase in abnor-
mal embryos is mainly due to an increase in mitotic segregation
errors in early embryonic cleavage divisions.
Patient selection does not explain observed differences in
aneuploidy rate
Although not significant (x2; P ¼ 0.097), a trend was observed
following mild stimulation towards a higher rate of drop out
before PGS analysis, since 27 out of 67 (40%) patients were
either lost before oocyte retrieval, fertilization or embryo
biopsy (Figure 2). After conventional stimulation, 11 out of
44 (25%) patients did not reach PGS analysis. The retrieval
of only a few oocytes after conventional stimulation has been
attributed to ovarian ageing (Beckers et al., 2002; de Boer
et al., 2002), and an age-dependent increase in chromosomal
abnormalities in oocytes has been reported (Hassold and
Hunt, 2001). It is possible that women with more advanced
ovarian aging undergoing mild stimulation were less likely to
meet the criteria for oocyte retrieval, thus creating a selection
bias for women with a reduced incidence of aneuploid
embryos. To exclude such a potential selection bias, female
age and two distinct markers for ovarian ageing (early follicu-
lar phase FSH and inhibin B levels) (Groome et al., 1996;
Creus et al., 2000) were retrospectively compared between
the patients who did and those patients who did not reach
PGS following mild stimulation. No differences were observed
in age (33.2 + 3.2 versus 32.3 + 3.4 years; P ¼ 0.31),
baseline serum levels of FSH (7.8 + 2.2 IU l21 versus
7.7 + 3.3 IU l21; P ¼ 0.93) or inhibin B (110 + 75 ng l21
versus 108 + 129 ng l21; P ¼ 0.96). Therefore, we find no
indications that women with more advanced ovarian ageing
showed a higher drop-out rate after mild ovarian stimulation.
However, it cannot be excluded that other mechanisms for
patient selection may be involved.

Figure 3. Distribution of number of oocytes retrieved per patient and relationship between oocyte number and percentage of abnormal embryos
generated following conventional (a and c) and mild ovarian stimulation for IVF (b and d).
Discussion
The introduction of GnRH antagonists allows ovarian
stimulation for IVF without disrupting early follicular phase
dynamics. In the present randomized trial, we compare the
effect of a mild stimulation approach to a conventional stimu-
lation regimen by assessing chromosomal competence of
embryos. We found that mild stimulation is associated with a
reduction in the number of oocytes retrieved and embryos gen-
erated. However, the proportion of chromosomally normal
embryos is significantly increased. Consequently, the number
of chromosomally competent embryos obtained per woman is
similar (around two), despite a significant reduction in the
total number of embryos in the mild stimulation group. In
addition, analysis of two cells per embryo suggests that the
increase in chromosomal abnormalities observed after conven-
tional stimulation, is mainly due to an increased incidence of
chromosomal mosaicism.
In the mild stimulation group, patients received lower doses
of exogenous FSH. Since no down-regulation of endogenous
Ovarian stimulation affects oocyte quality
Downloaded from https://academic.oup.com/humrep/article-abstract/22/4/980/697233 by guest on 14 June 2019
FSH production has taken place, serum FSH concentrations
on cycle Day 8 were shown to be equivalent to those observed
in conventional stimulation with a high dose of exogenous FSH
(Hohmann et al., 2003). The difference between the two
stimulation protocols involves both follicle recruitment and
selection. In the natural cycle, a synchronous cohort of follicles
gains gonadotrophin dependence due to the intercycle rise in
endogenous FSH and continues its development. The dominant
follicle is selected around the mid-follicular phase from this
pool of 20– 30 antral follicles. Decreasing FSH concentrations
are crucial for single dominant follicle selection (Zeleznik and
Hillier, 1984; Fauser and Van Heusden, 1997). In addition, the
dominant follicle suppresses subdominant follicles through
intraovarian mechanisms (Baker and Spears, 1999). In mild
stimulation, interference with decreasing FSH gives rise to
the development of multiple dominant follicles, whereas
follicle recruitment and the initial stages of selection remain
unaffected. In contrast, during conventional ovarian stimu-
lation, including pituitary down-regulation by GnRH agonist
985
E.B.Baart et al.

Table III. Outcomes after IVF and preimplantation genetic screening diagnosis following conventional or mild ovarian stimulation

Conventional stimulation Mild stimulation P* Difference (95% CI)

IVF characteristics
No. of patients 40 55a
Oocytes retrieved (n) 12.1 + 5.7 8.3 + 4.7 ,0.01 3.7 (1.6–5.9)
Fertilization rate (%) 57 + 28 55 + 30 0.81 1.5 (210–13)
Embryos (2pn) 6.8 + 5.0 4.7 + 3.9 0.03 2.0 (0.2–3.9)
Good quality embryo rateb (%) 35 + 29 51 + 40 0.04 217 (232– 1)
Diagnosis based on first cell biopsiedc
No. of patients 33 40
Embryos diagnosed 4.8 + 3.5 3.6 + 2.7 0.10 1.2 (20.2– 2.7)
Percentage of embryos diagnosed (%) 40 + 22 45 + 23 0.38 25 (215–6)
Abnormal embryos/embryos diagnosed (%) 63 + 28 45 + 35 0.016 19 (4– 34)
Diagnosis based on two cellsd
No. of patients 30 38

Downloaded from https://academic.oup.com/humrep/article-abstract/22/4/980/697233 by guest on 14 June 2019

Abnormal embryos/embryos diagnosed (%) 73 + 33 55 + 42 0.046 19 (0.3–36)
Mosaic embryos/embryos diagnosed (%) 65 + 37 37 + 39 0.004 28 (10–47)
Clinical outcome measures
Embryos/transfer 1.45 + 0.51 1.46 + 0.51
Ongoing pregnancy rate/started cycle (%) 7/41 (17) 12/63 (19)
Ongoing pregnancy rate/transfer (%) 7/31 (23) 12/35 (34)

Data are expressed on a per patient basis and are presented as mean and SD, unless otherwise stated.
* P-values are from a two-sample t-test.
a
One patient out of the 56 undergoing oocyte retrieval yielded no oocytes.
b
Embryos with normal morphology were defined as embryos with timely development, ,20% fragmentation, equally sized blastomeres and small or no
irregularities observed in the cytoplasm.
c
Diagnosis of normal or abnormal embryos was based on the FISH results of one cell. If two cells were available, the first cell biopsied was determined in
retrospect and used for diagnosis. Rates were calculated first per patient and then averaged.
d
Only embryos where two cells were available for diagnosis were taken into account. An embryo was considered abnormal if at least one of the two cells
showed an abnormal result.

co-treatment, natural follicle recruitment and selection is
completely overruled, allowing the non-discriminate growth
of many follicles at different developmental stages.
Following recruitment into the growing pool, the oocyte
expands from 35 to 120 mm in diameter, which represents a
100-fold increase in volume over a period of several months
(Gosden and Bownes, 1995). Oocyte growth and maturation
is interlinked with follicle development, and bi-directional
signalling occurs between oocytes and granulosa cells (Eppig,
2001). Oocytes have to achieve both nuclear and cytoplasmic
maturity in order to sustain the early stages of embryonic
development (Albertini et al., 2003). Recently, experimental
evidence in mice showed that disturbances in the complex
interplay of signals regulating folliculogenesis may alter the
late stages of oocyte growth, increasing the risk for chromo-
some malsegregation in subsequent meiotic divisions
(Hodges et al., 2002). These findings offer a rationale for our
findings of an increased proportion of chromosomally normal
embryos after mild ovarian stimulation. However, the possi-
bility that the different GnRH analogues directly influence
the chromosomal constitution of the embryos in this study
cannot be ruled out.
Interestingly, our results suggests that the increase in the pro-
portion of abnormal embryos was mainly due to an increase in
mitotic segregation errors, leading to mosaic embryos. The
embryonic genome does not become active until the eight
cell stage (Braude et al., 1988), until then the cell cycle
machinery is dependent on the protein and mRNA content of
the oocyte. Recently, a direct link has been established
between defects in the oocyte and an increased incidence in
mitotic segregation errors. An experimental mouse model
986
with an inactivated protein subunit of the meiotic synaptone-
mal complex (SYCP3) revealed not only an increased level
of segregation errors at the first meiotic division but also
showed a substantial increase in mitotic segregation errors
during the first embryo cleavage divisions (Yuan et al., 2002;
Lightfoot et al., 2006). More research into the developmental
potential of embryos with mitotic segregation errors is
needed to understand the significance of mosaicism in human
embryos. However, there are indications that the implantation
potential of embryos mosaic for trisomy 21 is reduced
(Katz-Jaffe et al., 2004).
Within the mild stimulation group, we also found that a
low oocyte yield is associated with a decrease in the
proportion of aneuploid embryos. A previous study showed
mild stimulation to result in high-quality embryos for trans-
fer, as indicated by good embryo morphology, and pregnancy
rates comparable to those following conventional ovarian
stimulation (Hohmann et al., 2003). Moreover, although no
pregnancies were obtained in women who had produced
four or less oocytes following the conventional protocol,
the majority of pregnancies obtained following mild
ovarian stimulation occurred in women where four or less
oocytes were retrieved. A low number of oocytes retrieved
after stimulation may, therefore, represent an appropriate
response to mild stimulation. In contrast, a similar low
response occurring after conventional ovarian stimulation is
indeed indicative of ovarian ageing (Beckers et al., 2002;
de Boer et al., 2002). Although few pregnancies were
achieved, the pregnancy rates we observed after PGS are
within the range reported by the ESHRE PGD consortium
(Harper et al., 2006).

Figure 4. Oocyte and embryo yield and embryos successfully
biopsied and diagnosed by fluorescent in-situ hybridization (FISH)
as chromosomally normal on the basis of FISH results from one cell
following conventional and mild stimulation. Values are expressed
on a per patient basis and are represented as mean. Bars indicate the
95% confidence interval. *P , 0.05, **P , 0.01.
Although implantation, ongoing pregnancy and ultimately
live birth are the most meaningful outcome measures, they
are only partially influenced by embryo quality and can only
be determined for the embryos transferred. In the current
study, PGS was used as a parameter for assessing embryo
quality. It revealed a significant effect of the ovarian
stimulation regimen on the chromosome segregation ability
of the resulting embryos. This observation supports the hypoth-
esis that only the follicle with the most competent oocyte is
selected during the natural cycle in the mono-ovulatory
human species. The present mild stimulation protocol rep-
resents less interference with ovarian physiology, which may
give rise to a higher proportion of developmentally competent
oocytes. This concept is also consistent with an extensive
analysis of historical data showing no significant improvement
of the pregnancy rate per oocytes retrieved using ovarian
stimulation when compared with IVF results in the early
1980s, when IVF was performed without ovarian stimulation
(Inge et al., 2005).
In conclusion, the present study shows, for the first time, that
mild ovarian stimulation results in fewer oocytes and a decreased
proportion of aneuploid and mosaic embryos. Obviously, our
findings need to be confirmed by other groups, as both treatment
strategies and PGS methodologies vary largely between centres.
However, based on the current findings, we would like to propose
that future ovarian stimulation strategies should not focus on
obtaining as many oocytes as possible. Instead, strategies
should aim at less interference with ovarian physiology, thus
minimizing embryo aneuploidy rate and facilitating selection of
the best quality embryo for transfer.
Ovarian stimulation affects oocyte quality
Acknowledgements
The authors like to thank D. Berks, Medical Center Rijnmond Zuid,
Rotterdam, The Netherlands, for his assistance in patient inclusion
and I. van den Berg, D. Bulkmans and L. Nekrui, Erasmus MC,
Rotterdam, The Netherlands for technical assistance and help with
collecting blood samples. Prof. A. Hsueh, Stanford University,
Stanford, USA and Dr P. de Boer, University Medical Centre St
Radboud, Nijmegen, The Netherlands are gratefully acknowledged
for critically reviewing the manuscript. This research was financially
supported by the Erasmus University (AIO) and the ‘Stichting
Voortplantingsgeneeskunde Rotterdam’.
References
Albertini DF, Sanfins A and Combelles CM (2003) Origins and manifestations
Downloaded from https://academic.oup.com/humrep/article-abstract/22/4/980/697233 by guest on 14 June 2019
of oocyte maturation competencies. Reprod Biomed Online 6,410– 415.
Baart EB, Martini E and Van Opstal D (2004a) Screening for aneuploidies of
ten different chromosomes in two rounds of FISH: a short and reliable
protocol. Prenat Diagn 24,955– 961.
Baart EB, Van Opstal D, Los FJ, Fauser BC and Martini E (2004b)
Fluorescence in situ hybridization analysis of two blastomeres from day 3
frozen-thawed embryos followed by analysis of the remaining embryo on
day 5. Hum Reprod 19,685– 693.
Baart EB, Martini E, van den Berg I, Macklon NS, Galjaard RJ, Fauser BC and
Van Opstal D (2006) Preimplantation genetic screening reveals a high
incidence of aneuploidy and mosaicism in embryos from young women
undergoing IVF. Hum Reprod 21,223–233.
Baker SJ and Spears N (1999) The role of intra-ovarian interactions in the
regulation of follicle dominance. Hum Reprod Update 5,153–165.
Beckers NG, Macklon NS, Eijkemans MJ and Fauser BC (2002) Women with
regular menstrual cycles and a poor response to ovarian hyperstimulation for
in vitro fertilization exhibit follicular phase characteristics suggestive of
ovarian aging. Fertil Steril 78,291–297.
Bielanska M, Tan SL and Ao A (2002) Chromosomal mosaicism throughout
human preimplantation development in vitro: incidence, type, and
relevance to embryo outcome. Hum Reprod 17,413– 419.
Boué J, Boué A and Lazar P (1975) Retrospective and prospective
epidemiological studies of 1500 karyotyped spontaneous human abortions.
Teratology 12,11–26.
Braude P, Bolton V and Moore S (1988) Human gene expression first occurs
between the four- and eight-cell stages of preimplantation development.
Nature 332,459– 461.
Champion MD and Hawley RS (2002) Playing for half the deck: the molecular
biology of meiosis. Nat Cell Biol 4 (Suppl.),s50–s56.
Cohen J and Munné S (2005) Comment 2 on Staessen et al. (2004). Two-cell
biopsy and PGD pregnancy outcome. Hum Reprod 20,2363– 2364.
Creus M, Penarrubia J, Fabregues F, Vidal E, Carmona F, Casamitjana R,
Vanrell JA and Balasch J (2000) Day 3 serum inhibin B and FSH and age
as predictors of assisted reproduction treatment outcome. Hum Reprod
15,2341– 2346.
de Boer EJ, den Tonkelaar I, te Velde ER, Burger CW, Klip H and van
Leeuwen FE (2002) A low number of retrieved oocytes at in vitro
fertilization treatment is predictive of early menopause. Fertil Steril
77,978–985.
Delhanty JD (1997) Chromosome analysis by FISH in human preimplantation
genetics. Hum Reprod 12,153– 155.
Delhanty JD, Griffin DK, Handyside AH, Harper J, Atkinson GH, Pieters MH
and Winston RM (1993) Detection of aneuploidy and chromosomal
mosaicism in human embryos during preimplantation sex determination
by fluorescent in situ hybridisation, (FISH). Hum Mol Genet 2,1183–1185.
Eppig JJ (2001) Oocyte control of ovarian follicular development and function
in mammals. Reproduction 122,829–838.
Evers JL (2002) Female subfertility. Lancet 360,151– 159.
Fauser BC and Van Heusden AM (1997) Manipulation of human ovarian
function: physiological concepts and clinical consequences. Endocr Rev
18,71–106.
Fauser BC and Devroey P (2005) Why is the clinical acceptance of
gonadotrophin-releasing hormone antagonist cotreatment during
ovarian hyperstimulation for in vitro fertilization so slow? Fertil Steril
83,1607–1611.
Fauser BC, Devroey P, Yen SS, Gosden R, Crowley WF Jr, Baird DT and
Bouchard P (1999) Minimal ovarian stimulation for IVF: appraisal of
potential benefits and drawbacks. Hum Reprod 14,2681–2686.
987
E.B.Baart et al.
Fauser BC, Devroey P and Macklon NS (2005) Multiple birth resulting from
ovarian stimulation for subfertility treatment. Lancet 365,1807– 1816.
Foote RH and Carney EW (1988) Factors limiting reproductive efficiency in
selected laboratory animals. Ann NY Acad Sci 541,683– 696.
Ghosh D, Stewart DR, Nayak NR, Lasley BL, Overstreet JW, Hendrickx AG
and Sengupta J (1997) Serum concentrations of oestradiol-17beta,
progesterone, relaxin and chorionic gonadotrophin during blastocyst
implantation in natural pregnancy cycle and in embryo transfer cycle in
the rhesus monkey. Hum Reprod 12,914– 920.
Gianaroli L, Magli MC, Fiorentino F, Baldi M and Ferraretti AP (2003)
Clinical value of preimplantation genetic diagnosis. Placenta 24(Suppl B),
pS77–S83.
Gosden RG and Bownes M (1995) Cellular and molecular aspects of oocyte
development. In Grudzinskas JG and Yovich JL (eds) Gametes—The
Oocyte. Cambridge Reviews in Human Reproduction, Cambridge
University Press, Cambridge, pp.23–53.
Groome NP, Illingworth PJ, O’Brien M, Pai R, Rodger FE, Mather JP and
McNeilly AS (1996) Measurement of dimeric inhibin B throughout the
human menstrual cycle. J Clin Endocrinol Metab 81,1401–1405.
Harper JC, Boelaert K, Geraedts J, Harton G, Kearns WG, Moutou C,
Muntjewerff N, Repping S, SenGupta S, Scriven PN, et al. (2006)
ESHRE PGD Consortium data collection V: Cycles from January to
December 2002 with pregnancy follow-up to October 2003. Hum Reprod
21,3– 21.
Hassold T and Hunt P (2001) To err (meiotically) is human: the genesis of
human aneuploidy. Nat Rev Genet 2,280– 291.
Hodges CA, Ilagan A, Jennings D, Keri R, Nilson J and Hunt PA (2002)
Experimental evidence that changes in oocyte growth influence meiotic
chromosome segregation. Hum Reprod 17,1171 –1180.
Hohmann FP, Macklon NS and Fauser BC (2003) A randomized comparison
of two ovarian stimulation protocols with gonadotrophin-releasing hormone
(GnRH) antagonist cotreatment for in vitro fertilization commencing
recombinant follicle-stimulating hormone on cycle day 2 or 5 with
the standard long GnRH agonist protocol. J Clin Endocrinol Metab
88,166– 173.
Huisman GJ, Fauser BC, Eijkemans MJ and Pieters MH (2000) Implantation
rates after in vitro fertilization and transfer of a maximum of two
embryos that have undergone three to five days of culture. Fertil Steril
73,117– 122.
Inge GB, Brinsden PR and Elder KT (2005) Oocyte number per live birth in
IVF: were Steptoe and Edwards less wasteful? Hum Reprod 20,588–592.
Katz-Jaffe MG, Trounson AO and Cram DS (2004) Mitotic errors in
chromosome 21 of human preimplantation embryos are associated with
non-viability. Mol Hum Reprod 10,143–147.
Katz-Jaffe MG, Trounson AO and Cram DS (2005) Chromosome 21 mosaic
human preimplantation embryos predominantly arise from diploid
conceptions. Fertil Steril 84,634–643.
Lightfoot DA, Kouznetsova A, Mahdy E, Wilbertz J and Hoog C (2006) The
fate of mosaic aneuploid embryos during mouse development. Dev Biol
289,384–394.
Macklon NS, Stouffer RL, Giudice LC and Fauser BC (2006) The science
behind 25 years of ovarian stimulation for IVF. Endocr Rev 27,170–207.
988
Munné S (2006) Chromosome abnormalities and their relationship to
morphology and development of human embryos. Reprod Biomed Online
12,234–253.
Munné S, Magli C, Adler A, Wright G, de Boer K, Mortimer D, Tucker M,
Cohen J and Gianaroli L (1997) Treatment-related chromosome
abnormalities in human embryos. Hum Reprod 12,780–784.
Munné S, Marquez C, Magli C, Morton P and Morrison L (1998)
Scoring criteria for preimplantation genetic diagnosis of numerical
abnormalities for chromosomes X, Y, 13, 16, 18 and 21. Mol Hum Reprod
4,863– 870.
Munné S, Sandalinas M, Escudero T, Velilla E, Walmsley R, Sadowy S, Cohen
J and Sable D (2003) Improved implantation after preimplantation genetic
diagnosis of aneuploidy. Reprod Biomed Online 7,91–97.
Norwitz ER, Schust DJ and Fisher SJ (2001) Implantation and the survival of
early pregnancy. N Engl J Med 345,1400–1408.
Pehlivan T, Rubio C, Rodrigo L, Romero J, Remohi J, Simon C and Pellicer A
(2003) Impact of preimplantation genetic diagnosis on IVF outcome in
Downloaded from https://academic.oup.com/humrep/article-abstract/22/4/980/697233 by guest on 14 June 2019
implantation failure patients. Reprod Biomed Online 6,232–237.
Platteau P, Staessen C, Michiels A, Van Steirteghem A, Liebaers I and
Devroey P (2005) Preimplantation genetic diagnosis for aneuploidy
screening in patients with unexplained recurrent miscarriages. Fertil Steril
83,393–397.
Pocock SJ (1977) Group sequential methods in the design and analysis of
clinical trials, Biometrika 64,191– 199
Staessen C, Platteau P, Van Assche E, Michiels A, Tournaye H, Camus M,
Devroey P, Liebaers I and Van Steirteghem A (2004) Comparison of
blastocyst transfer with or without preimplantation genetic diagnosis for
aneuploidy screening in couples with advanced maternal age: a
prospective randomized controlled trial. Hum Reprod 19,2849 –2858.
Stevens VC (1997) Some reproductive studies in the baboon. Hum Reprod
Update 3,533–540.
Taylor A (2003) ABC of subfertility: extent of the problem. BMJ 327,434–436.
Thornhill AR, deDie-Smulders CE, Geraedts JP, Harper JC, Harton GL, Lavery
SA, Moutou C, Robinson MD, Schmutzler AG, Scriven PN, et al. (2005)
ESHRE PGD Consortium ‘Best practice guidelines for clinical
preimplantation genetic diagnosis (PGD) and preimplantation genetic
screening (PGS)’. Hum Reprod 20,35–48.
Vorsanova SG, Kolotii AD, Iourov IY, Monakhov VV, Kirillova EA, Soloviev
IV and Yurov YB (2005) Evidence for high frequency of chromosomal
mosaicism in spontaneous abortions revealed by interphase FISH analysis.
J Histochem Cytochem 53,375– 380.
Wilcox AJ, Weinberg CR, O’Connor JF, Baird DD, Schlatterer JP, Canfield
RE, Armstrong EG and Nisula BC (1988) Incidence of early loss of
pregnancy. N Engl J Med 319,189– 194.
Yuan L, Liu JG, Hoja MR, Wilbertz J, Nordqvist K and Hoog C (2002) Female
germ cell aneuploidy and embryo death in mice lacking the meiosis-specific
protein SCP3. Science 296,1115–1118.
Zeleznik AJ and Hillier SG (1984) The role of gonadotrophins in the selection
of the preovulatory follicle. Clin Obstet Gynecol 27,927– 940.
Submitted on August 23, 2006; resubmitted on October 17, 2006; resubmitted
on November 8, 2006; accepted on November 29, 2006